A Chinese-language information service for current AI news, selected stories, hot topics, daily reports, and papers. It retrieves fresh results from AIHOT’s public read-only API rather than relying on remembered information.
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Review a branch or pull request for correctness, lifecycle, security, compatibility, and missing behavior using the target repository's own instructions and live base/head rather than DeepSeek Harness-specific commands.
Find evidence-backed opportunities to remove dead, duplicated, speculative, over-built, or hand-rolled code without treating complexity, tests, or unused-looking symbols as proof by themselves.
Write, review, or trim technical prose in Markdown, JSDoc, comments, prompts, diagnostics, and visible strings while preserving complete contracts and removing repetition or authoring-session narration.
Audit or rewrite repository prose that exposes authoring-session reasoning, review history, dead draft references, change narration, or unsupported planning residue while preserving every durable technical fact.
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when the user wants to stress-test a plan, get grilled on a design, or says "grill me".
First dig out what actually matters in a project or system from real sources, then teach it one small point at a time, pausing for confirmation after each. Use when the user wants to genuinely understand a codebase/system (not get a summary), is preparing to explain their own work (e.g. interview prep on a resume)…
Resume structure, achievement bullet formulas, ATS optimization, and job-targeted tailoring for software engineers. Use when reviewing resumes, crafting achievement bullets, extracting keywords from job descriptions, or tailoring content for specific roles.
Use when implementing any feature, bug fix, refactor, or behavior change. Write a failing behavioral test first, verify the failure, then write the minimum production code needed to pass.
Conducts multi-axis code review. Use before merging any change. Use when reviewing code written by yourself, another agent, or a human. Use when you need to assess code quality across multiple dimensions before it enters the main branch.
A writing assistant for academic papers and related documents, such as theses, conference papers, literature reviews, and research plans. It helps shape the question, structure the argument, organize evidence, and revise the text.
A personal finance planning guide that organizes income, spending, debt, savings, risks, and future goals. It provides educational planning, not licensed financial, tax, legal, or investment advice.
A skill that specializes in conducting thorough code reviews. Provide task description, MR summary, and diff command (DO NOT provide specific code changes. Provide command to get diffs like 'git add . && git diff HEAD --relative -- ".go" ".yaml"') to begin review.
Use when creating, packaging, reviewing, or improving an Agent Skill for Codex, Claude, Trae, or another SKILL.md-compatible runtime.
778 3d agoC34 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: